The world is in turmoil. We are seeing frustration, arguments, polarization, extremism and completely irrational behavior everywhere we look.
Congress is in gridlock. Terrorism has become pervasive. People are angry and argumentative. Hostilities and uprisings are occurring all around the world. Wars are a sign of problems that have not been solved that have thus been reframed as how can we intimidate or eliminate those others who disagree.
And if we understand why this is happening, we should not be the least bit surprised!!
When people are unable to see the problems being solved that adversely affect them, they become frustrated and irrational.
With advances in technologies and globalism, everything and everyone is so intertwined today that we can now create systems that are so complex that we are unable to predict the consequences of our actions. This creates even more problems that we cannot solve and makes the world even more complex and chaotic. This process is infectious.
Might we use the computer to help us solve some of these problems that we previously could not solve that have created so much mayhem?
Humans are inherently limited to a very few cause-and-effects they can deal with at one time to solve problems.
George A Miller’s research paper in 1956 set the limit on how many independent items we can keep in mind at one time as 7 plus or minus 2. The number of cause-and-effects we can keep in mind and also logically connect them to find their implications is probably well less than Miller’s limit.
Can we solve these problems and reduce some of the chaos?
Yes, we can use a computer program that can put together many more cause-and-effects to find their implications and solve problems than humans are capable of.
Using the Explainer computer program to deal with the problems that lead to the gridlocks in Congress required dealing with the order of forty or more cause-and-effects. This is far beyond our human capabilities. The Explainer presents us with the opportunity to escape some of the dilemmas and chaos we are facing today.
Some of the problems that have Congress in gridlock that we, the Explainer and I, have been able to shed light on are:
- Finding the likely causes of the economic crisis and widening wealth gap and revealing a fundamental weakness in the capitalist system and how to resolve this.
- Finding a way of improving the economy by increasing demand.
- Reframing the problem of how fast to decrease the deficit so it might be more easily solved.
The Explainer helps solve problems that can be stated as: Given a behavior and a set of cause-and-effects that include those pertinent to the problem, what causes that behavior (abduction). We can trace backwards through the cause-and-effects to find explanations for the behavior. Explanations are Boolean functions of assumptions using ANDs, ORs, and NOTs. You might consider the assumptions as possibilities to be considered to solve the problem.
Then given any one of the explanations, we can trace forward through the cause-and-effects to find what that explanation would predict (deduction).
Now that we have the assumptions that make up the explanations for the behavior, we can ask about the significance of those assumptions. We may already have facts that establish that certain assumptions are true. Or we may not know whether an assumption is true, in which case we may investigate further to determine whether it is true. Or we may be uncertain as to whether the assumptions are true and thus assign probabilities to the likelihoods that they are true.
I might make this clearer if we don’t get hung up on the word ‘assumptions’. Consider instead the word ‘possibilities’, the various possibilities that might explain the behavior. Then we can say that we start with a behavior and consider the various possibilities that could explain that behavior. These possibilities might be obvious as a consequence of our experience, or they may have to be worked out by tracing through a system of cause-and-effects as is done by the Explainer.
The Explainer handles circuits in the cause-and-effect circuits, which can be vital to solving many problems. If A causes B, and B causes C, and C causes A, this would be a cause-and-effect circuit.
But neither Expert Systems nor Bayesian Networks can handle cause-and-effect circuits. Causal Loop Diagrams do not reveal the many implications of circuits that are revealed by the Explainer
The Explainer can also be used to design systems by specifying their behavior, explaining the causes of that behavior, and using one of the explanations to implement the assumptions, i.e. possibilities, as the design.
In solving problems, the Explainer can show what further information is required and given that information find a resolution to the problem. It can use this capability to deal with problems such as:
- Medical diagnosis
- Finding faults in systems
- Analyzing crimes
- Analyzing the causes of historic and archeological events
- Even seeing how terrorists can plan to cause us harm so that knowing their plans we can intervene to thwart them.
I would like to see a website developed where people who have some familiarity with a problem and the Explainer method could collaborate to solve problems. Those who have the ability to solve problems that others are unable to solve gain power over the others.
Could this power be used to allow people to solve the problems that Congress and the government have been unable to solve? This would give us a government that works bottom-up from the people to a government that serves the people. This power would let the people compete with the power of hidden sources of money that work from the top down to benefit a small minority. This could help the people restore some of the democracy that we have been steadily losing.
Let me give several examples of problems that bewilder Congress and how we can cast new light on them.
Consider Congresses’ argument over how fast to reduce the deficit. Some want to reduce the deficit as quickly as possible. That would solve today’s short-term problems. Others wish to invest in education, infrastructure, and research now to create a more prosperous tomorrow. But this would raise the deficit today. There are several mechanisms whereby reducing the deficit would affect tomorrow’s prosperity. Reducing the deficit today will leave a lower deficit to be faced by people tomorrow. But it would also force out investment in education, infrastructure, and research that while increasing the deficit today will increase the capability of those who can pay off the deficit tomorrow. Perhaps we should reframe the issue as: How do we improve our prosperity tomorrow at the least cost to prosperity today. This may encourage those who want to decrease the deficit faster today to think of in the context of its implications for tomorrow.
Another more complex problem is finding the causes of the economic crisis and the widening wealth gap, and then figure out what to do about it. The Explainer traces this back to the process where some people in a negotiation possess the knowledge needed to evaluate their own self-interests, giving them an undue advantage over those who do not possess the knowledge needed to evaluate their own self-interests. Think of the subprime mortgage negotiations.
It may be hard to understand how the Explainer arrived at its conclusions so that people would feel comfortable believing it. But it often casts new light on the problem, suggesting something we already know that we had not considered. Considering what we had overlooked would make it easier to believe, and to get the appropriate people to implement the Explainer’s proposed solutions.
Here is something we may have overlooked while considering the problem of how fast to decrease the deficit: Adam Smith in his 1776 book The Wealth of Nations told us that if everyone negotiates in his own self-interests, the result would be an economy that allocates resources efficiently and allows each to gain from the economy in measure to what he contributes to the economy. This has been the basis of the capitalist system. The capitalist system often works well, but sometimes causes us much trouble. An example is the economic crisis and widening wealth gap.
There is an assumption in Smith’s depiction of capitalism that is usually overlooked, and sometimes used maliciously. The assumption is that each of the parties in a negotiation has equal knowledge to evaluate his self-interests. This weakness in the capitalist system may be overcome by the functioning of the Consumer Financial Protection Bureau, if Congress allows it to function properly.
Another interesting problem is how the lack of demand may be hurting our economy. An Explainer analysis of this problem involves dealing with a cause-and-effect circuit. (Other methods that deal with cause-and-effects such as Rule Based Expert Systems or Bayesian Networks do not deal with cause-and-effect circuits. Causal Loops do not deal with circuits sufficiently for our purposes.)
Let’s trace this circuit:
- Businesses will not increase their capacity to produce more goods and services if there is no demand for them.
- People will not increase their demand if they do not have the money to satisfy their demands.
- People would have that money to satisfy their demands if businesses were to hire them at good salaries.
- Hiring more people would increase businesses’ capacity to meet the demand.
This completes the circuit.
Circuits generally have two solutions. One solution causes a cycling up, while the other solution causes a cycling down. Which cycle prevails is determined by external causes that can drive one or the other cycle.
If there is not an external cause to drive one of the solutions, it can cycle equally well either up or down, i.e. it can float. This suggests that by driving this cycle by increasing the minimum wage would cause the system to cycle up, creating more demand, more business, and people to be paid more. This would be a benefit to the overall economy.
But further analysis reveals a difficulty. Much of our production today is done by automated equipment rather than by employees. But automated equipment does not produce demand nor does it pay taxes as employees would. This suggests that we should tax automated equipment and provide tax credits for businesses that create more employment. This would be consistent with Thomas Piketty’s proposal in his book Capital in the Twenty First Century that wealth should be taxed. Except for property taxes, we do not currently tax wealth. But a tax on automated equipment would be a tax on wealth. Such a tax on automated equipment and a tax credit to businesses for hiring more employees might be in order. It might also reduce the effective wages in the U.S. that are currently causing businesses to offshore their production or base their business oversees where they can pay lower taxes. This may call for more consideration.
How it works
If you are interested in further detail, you may wish to read the following.
An Explainer problem begins with the input of a matrix showing the pertinent cause-and-effects. Figure 1 shows the input for the problem of how fast to reduce the deficit. It is presented in the form of a matrix of cause-and-effects. The columns are the causes and the rows are the effects. The causes in the column and the effects in the rows are the same because an effect can become the cause of another effect. Their labels appear to the left of the matrix.
The effects in the matrix are so ordered that the cause appears earlier in the matrix than its effect except for the circuits, which appear in blocks on the diagonal. A mark above the diagonal will only occur within a square block shown on the diagonal.
The black marks in the diagonal only show where the cause and the effect in the matrix are the same. They are for reference only to help traversing the matrix.
A variable is associated with each effect. We are interested in whether the value of that variable moves Up or Down. This is often sufficient to determine what actions to take.
A mark in the matrix shows that the cause from its column produces the effect in its row. For example, the S in column 2 row 6 indicates that an increase in ‘2. Automation’ causes an increase in ‘6. Joblessness / Low payrolls’. Conversely, a reduction in ‘2. Automation’ would cause a reduction in ‘6. Joblessness / Low payrolls’.
An S in the matrix represents a cause-and-effect such that the value associated with the effect moves in the same direction as the value associated with the cause. An O represents a cause-and-effect where the values of the cause and its effect move in the opposite direction. For example, the O in column 5 row 6 indicates that ‘5. Stimulus and Quantitative easing’ has the opposite effect on ‘6. Joblessness / Low payrolls, i.e. ‘6. Joblessness /Low payrolls’ decreases. Or if you prefer, ‘Jobs and Payrolls’ increase.
The matrix shows the direct cause-and-effects. But an effect can be the cause of another effect, making it possible to string together paths of cause-and-effects to show how one effect indirectly affects another. So using the matrix, we can find cause-and-effect paths. A single cause-and-effect is a direct path. A string of cause-and-effects where intermediate effects are the causes of other effects is an indirect path.
We can put together cause-and-effects to form paths, or what we sometimes call mechanisms. For example, looking at the O in column 5 row 6, and the S in column 6 row 18, and the S in column 18 row 19, we see the following path: When ’5. Stimulus and Quantitative easing’ moves Up, it causes ‘6. Joblessness / Low payrolls’ to move down, and that causes ’18. Spending on education, infrastructure, and research’ to move Up, that causes ’19. Future Prosperity and Competitive advantage over other countries’ to move Up. Tracing a path is like going down a stair step.
The S’s and O’s in the matrix show changes. Given that the first effect in a path is assigned a U, the path then shows how the other effects in the path move Up or Down accordingly. If that first effect were assigned a D, the assignments of U or D of the other effects in the path would be reversed.
The matrix is partitioned, meaning that all the marks that do not occur within circuits appear on one side of the diagonal, while the cause-and-effects that form circuits occur within the square blocks on the diagonal.
A block is a maximal subset of the effects such that there is a path, direct or indirect, from every effect in the block to every other effect within the same block. It is maximal because this would not be the case if any additional effect where added to the block. Each circuit occurs in only one block. Blocks may contain more than one circuit.
Note that if we were to make a mark in the block for every indirect connection between effects in the block, the block would be filled with marks. This is the consequence of the definition of blocks. So we don’t do that.
Figure 1: Input Cause-and-Effects shown as a partitioned matrix
We can put together cause-and-effects to form paths, or what we sometimes call mechanisms. For example, looking at the O in column 5 row 6 and the S in column 6 row 18 and the S in column 18 row 19, we see the following path: If‘’5. Stimulus and Quantitative easing’ moves Up, it causes ‘6. Joblessness / Low payrolls’ to move down, or in other words it causes ‘More jobs and Higher payrolls, that causes’18. Spending on education, infrastructure, and research’ to move Up, that causes ’19. Future Prosperity and Competitive advantage over other countries’ to move Up.
Cause-and-effects are transitive. That is: If ‘A causes B’ and ‘B causes C’, we can eliminate the B to get A causes C. We can find the net effect of indirect paths using elimination by substitution just as we do for solving simultaneous equations to eliminate the intermediate effects to show just the effect of the beginning of the path on the effect at the ending of an indirect path. So this indirect path amounts to: if ‘5. Stimulus and Quantitative easing’ moves Up, it causes ’19. Future Prosperity and Competitive advantage over other countries’ to also move Up.
The matrix in Figure 1 is partitioned, meaning that all the marks that do not occur within circuits appear on one side of the diagonal, while the cause-and-effects that form circuits occur within the square blocks straddling the diagonal.
The matrix shows direct paths involving just one cause-and-effect. For example, the S in column ‘7. Decrease in deficit burden on future generations’ causes an increase in ’19. Future Prosperity and Competitive advantage over other countries’.
In solving the deficit reduction crisis, we find three paths through the matrix that represent mechanisms whereby ‘3. Severe deficit reduction’ causes ’19 Future Prosperity and Competitive advantage over other countries’ to move either Up or move Down.
Mechanism 1: An increase in ‘4. Severe deficit reduction’ causes ‘8. Decrease in deficit burden on future generations’, which causes an increase in ’19. Future prosperity and Competitive advantage over other countries’.
Mechanism 2: An increase in ‘4. Severe deficit reduction’ causes a decrease in ‘18. Spending on education, infrastructure, and research’ that causes a decrease in ’19. Future prosperity and Competitive advantage over other countries’.
Mechanism 3: An increase in ‘4. Severe deficit reduction’ causes ‘8. Current austerity’ to move Up, that causes ‘9. High financial uncertainty about future’ to move Up, that causes ’10. Big businesses have cash hoard’ to move Up because they want to protect themselves against future financial uncertainty, that causes ’11. Tax increases on wealthy’ to move Up, that causes’12. Tax decrease on rest’ to move Up, that causes ’13 Increase in disposable income’ to move Up, that causes ’18. Spending on education, infrastructure, and research’ to move Up, that causes’19. Future prosperity and Competitive advantage over other countries’ to move Up. These mechanisms were found by using the Explainer program.
But a closer look at mechanism 3 shows a false intermediate conclusion that forces us to go back to look for errors that we might have made in some of cause-and-effects we assumed. In Thomas Piketty’s Capital in the Twenty First Century he suggested that countries should tax wealth. But, in the U.S., we do not tax wealth, except for property taxes. However, above we did suggest a tax on automated equipment, which is a tax on wealth. So making this analysis using cause-and-effects helps us catch and correct errors in the cause-and-effects we had assumed. Once we correct the errors in the cause-and-effects, we try again. This can be an iterative process.
The Explainer can be used to look at most any problem that can be framed as: Given a behavior and a set of cause-and-effects that includes those that are pertinent to the problem, what are the explanations involving assumptions that if true would explain that behavior?
The Explainer begins with the behavior to be explained and traces backward through the cause-and-effects to find the assumptions that comprise the explanations for that behavior. Each explanation is made up of functions of the assumptions using ANDs, ORs, and NOTs. There may be many plausible explanations for a given behavior.
An assumption can have one of several states. If all the assumptions of an explanation are true, i.e. they are known to be facts, then we can say that that explanation would be a plausible explanation for that behavior. If there are assumptions that are not yet supported by facts, they present a challenge to determine whether they are true. This may require more evidence or experiments. Or we may not have sufficient information to determine whether an assumption is true. So we use a probability of its being true.
If an explanation predicts behaviors that are known not to be true, that explanation must be rejected.
In the following example, the likelihood that an assumption is true plays a vital role. Consider the problem faced by a police officer confronting someone who shows signs of possible hostility. The police officer’s problem is how does he protect himself from that possible hostility while doing minimum harm to the person he is confronting? This problem could adversely affect him without his being able to see an immediate solution because of his lack of information. His training may help him obtain some of the information he needs quickly enough to avoid a catastrophe. But if not, he may become frustrated and inclined to act irrationally. Lives are at stake because each party does not understand the intentions of the other to avoid tragedy.
A similar situation may exist between nations, often leading to war.
The Explainer has also allowed us to deal which such problems as:
- How to find the causes of faults in systems,
- Given a patient’s medical symptoms, what are the possible diagnoses and what tests would be required to obtain the correct diagnosis, and what tests would not be required?
- Given the evidence concerning a crime, what further evidence might be needed, and with that evidence, determine who is the culprit.
- Given the desired behavior of a system, turn the assumptions for one of the explanations for that behavior into a design that would achieve that behavior.
The Explainer, if accepted, could be a very useful tool. Unfortunately, most people cannot conceive of how such a tool could be developed, and thus are inclined not to believe in it even when it is demonstrated to them. They suspect the person who claims to have developed the method must have made some mistake or has tricked them.
So now, we need to focus effort on getting people to consider this method more carefully and when they are satisfied, use it to solve those problems. Perhaps we can help create a world that is a little bit less frustrated and more rational and congenial.